Infectious disease specialist Jennifer Babik, MD, PhD, began her talk on curbside consults by explaining when not to ask an ID doctor for one.
“My personal opinion is you shouldn't do curbsides for [Staphylococcus] aureus bacteremia. A study last year found that curbsides were associated with a greater than 2-fold increase in mortality compared to bedside consults,” said Dr. Babik, an assistant clinical professor in the department of infectious diseases at the University of California, San Francisco (UCSF), speaking at UCSF's Management of the Hospitalized Patient conference last October.
Curbsides are appropriate for hypothetical, straightforward clinical questions, such as “What is the duration of therapy for a given disease?” she said. The questions shouldn't be so simple that the answer could be easily looked up or so complicated that the answer involves nuanced clinical judgment or a lot of data interpretation.
“We tell our infectious disease fellows that if you can't give the answer without looking it up, you should do a formal consult. If you can, and you trust the data source, it's OK to do a curbside,” she said.
Curbsides are not only permissible but necessary in many cases. Some infectious disease departments provide more than 10 curbsides per day on average, accounting for about 20% of their clinical work, according to Dr. Babik. “It would be impossible in most practices to convert all curbsides into formal consults,” she said.
Prior to her talk, Dr. Babik polled her UCSF colleagues about the most common questions they are asked during ID curbside consults. Interpreting urinalysis results was at the top of the list.
ASB or UTI?
Asymptomatic bacteriuria (ASB) is common in inpatients, and hospitalists often wonder if and how to treat it, Dr. Babik said.
ASB is defined as at least 105 bacteria from a single specimen for men and at least 105 bacteria on 2 separate specimens for women. In catheterized patients, the criterion is at least 102 bacteria from a single specimen for men or women. To be diagnosed with ASB, patients also shouldn't have any symptoms or signs related to urinary tract infection (UTI).
Patients who have had catheters for 30 days or longer have ASB 100% of the time; those who have had catheters for less than 30 days have it 9% to 23% of the time, Dr. Babik said. Other patients in whom ASB is common are those on hemodialysis (28%) and those with spinal cord injuries (23% to 89%) or diabetes (9% to 27%). In the subset of hospitalized patients with a urine culture positive for bacteria, about 90% have ASB, she said.
Most patients with ASB don't require treatment, Dr. Babik said. “Treatment doesn't reduce the risk of symptomatic UTI or later ASB and isn't associated with better outcomes in the long term,” she said. “This has been studied in premenopausal women, diabetic women, patients with spinal cord injuries, patients with catheters of short and long duration, and elderly patients.”
There are a few exceptions, however. Pregnant women should be treated because doing so decreases their risk of pyelonephritis, premature delivery, and low birth weight. Treatment also is advised in patients who are undergoing genitourinary procedures with expected mucosal bleeding, since they have high rates of postprocedure bacteremia (60%) and sepsis (up to 10%), and trials show treatment reduces these complications.
Immunosuppressed and transplant patients may also be candidates for treatment, although there aren't much data on this, and most transplant guidelines don't address it, Dr. Babik said. “Many centers do treat ASB in renal transplant patients in the first 3 months, and at my center, we treat neutropenic patients because of their risk for invasive disease,” she said.
While many patients with ASB don't need to be treated, it's difficult for physicians to ignore a positive urine culture, Dr. Babik said. “A lot of people like to give these patients antibiotics ‘just to be on the safe side,’” she said.
However, a proof of concept study published in 2014 in Clinical Infectious Diseases supports a hands-off approach, she added. In it, a center stopped reporting in their electronic medical record the positive urine cultures of noncatheterized inpatients, which reduced the percentage of those with ASB who were treated to 12% from 48%. There were no resulting untreated UTIs, and no patients developed sepsis. “This is probably my favorite study within the last year,” Dr. Babik said.
Part of the nervousness about not treating ASB is the fear that it may, in fact, be a UTI. Urinalysis is not helpful here, since pyuria is very common in patients with ASB, particularly those with catheters, where the positive predictive value of pyuria is only 32% to 36% for catheterized patients.
The same organisms cause ASBs and UTIs, so the specific microbiology isn't much use, either, Dr. Babik added. “Basically, to diagnose a UTI, the patient must have signs and symptoms compatible with a UTI, and no other source for infection—i.e., a diagnosis of exclusion,” she said.
Another common curbside question is whether Candida in the urine can be ignored, Dr. Babik said. Most of the time, it can.
“Candiduria is very common in patients with catheters and is usually asymptomatic. Just changing the Foley catheter can eliminate candiduria in 20% to 40% of patients,” she said.
Patients in whom candiduria should be treated are those who have symptoms of a UTI (defined similarly as in bacterial UTI) and those at high risk of dissemination—that is, neutropenic patients and those undergoing urologic procedures. Pregnant patients should also be treated.
First-line treatment is oral fluconazole, 200 to 400 mg daily for 14 days. Other treatment options have poor efficacy or a poor side-effect profile, including flucytosine, conventional amphotericin B, and amphotericin B bladder washes. Amphotericin B bladder washes do resolve candiduria in more than 90% of cases, but patients often relapse, she said.
Dr. Babik also briefly discussed Ebola, as this had been receiving a lot of attention at the time of the October 2014 conference. However, most questions about Ebola should be a formal consult, she noted.
Physicians should suspect Ebola if the patient has risk factors for exposure in the past 21 days (the incubation period), along with compatible symptoms such as fever, headache, weakness, muscle pain, diarrhea, vomiting, abdominal pain, and unusual bleeding or bruising.
“Ebola starts as a nonspecific febrile illness, and while before this outbreak we tended to think of hemorrhage with Ebola, bleeding manifestations are actually uncommon,” Dr. Babik said.
Polymerase chain reaction (PCR) testing for Ebola can be falsely negative in the first 3 days after symptoms, she noted, so retesting if there is a suspicion is important during this window.
The incubation period for Ebola is 2 to 21 days, while 8 to 10 days is the average. The nonspecific febrile illness phase—the first phase—lasts 3 to 4 days and resembles any other influenza-like illness. Then, the patient enters a 5- to 8-day gastrointestinal phase, in which he or she can have severe, watery diarrhea.
“Patients can lose 4 to 12 liters of stool a day, so keeping on top of these losses is a critical piece of management. You have to watch for severe hypovolemia and severe electrolyte abnormalities, even though patients may get very edematous,” she said.
Patients who survive usually begin to improve by day 6; those with fatal cases usually die between 6 and 16 days from symptom onset due to organ failure or shock, she said.
Infection control when treating Ebola is “a whole separate talk,” she said. The main points from CDC guidance on this are that health care workers should have repeated training on how to put on and take off personal protective equipment (PPE); no skin should be exposed; and all steps of PPE donning and doffing should be monitored and supervised by a trained observer, she noted.